Appendix for Mating systems and sex-biased disease

1 When does activity during the mating season start and end for each sex? (Figure 2, Top; Models 1-2)

1.1 Model 1: First, test effect of sex on date of first annual detection (i.e., start of swarm).

m1 = lmer(mindate2~sex + (1|site.wy2) + (1|pit_id),
              control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)), 
              data = df.m1); summary(m1)

Linear mixed model fit by REML ['lmerMod']
Formula: mindate2 ~ sex + (1 | site.wy2) + (1 | pit_id)
   Data: df.m1
Control: lmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))

REML criterion at convergence: 1119.1

Scaled residuals: 
    Min      1Q  Median      3Q     Max 
-3.5190 -0.5320 -0.0346  0.4860  3.9664 

Random effects:
 Groups   Name        Variance Std.Dev.
 pit_id   (Intercept) 0.19965  0.4468  
 site.wy2 (Intercept) 0.01754  0.1324  
 Residual             0.19144  0.4375  
Number of obs: 610, groups:  pit_id, 433; site.wy2, 7

Fixed effects:
            Estimate Std. Error t value
(Intercept)   1.0363     0.0599  17.301
sexF          0.2489     0.0766   3.249

Correlation of Fixed Effects:
     (Intr)
sexF -0.221

• Result: Females start autumn activity at hibernacula later than males.

1.2 Model 2: Then, test effect of sex on date of last annual detection (i.e., end of swarm).

m2 = lmer(maxdate2~sex + (1|site.wy2) + (1|pit_id),
              control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)),
              data = df.m2); summary(m2)

Linear mixed model fit by REML ['lmerMod']
Formula: maxdate2 ~ sex + (1 | site.wy2) + (1 | pit_id)
   Data: df.m2
Control: lmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))

REML criterion at convergence: 1672.9

Scaled residuals: 
     Min       1Q   Median       3Q      Max 
-2.90593 -0.39386  0.05173  0.51876  2.39627 

Random effects:
 Groups   Name        Variance Std.Dev.
 pit_id   (Intercept) 0.15645  0.3955  
 site.wy2 (Intercept) 0.07805  0.2794  
 Residual             0.20263  0.4501  
Number of obs: 948, groups:  pit_id, 609; site.wy2, 10

Fixed effects:
            Estimate Std. Error t value
(Intercept)  2.71097    0.09313  29.111
sexF        -0.38722    0.06159  -6.287

Correlation of Fixed Effects:
     (Intr)
sexF -0.095

• Result: Females end autumn activity at the hibernacula later than males.

1.3 Plot Figure 2, Top. Results of dates of mating activity analyses (Models 1-2)

• Top panel plots data and model coefficients that determined the active period
• Bottom panel visualizes peaks in activity by sex

print(p.fig2)

• Key findings: Females have a shorter active period, male activity fully encompasses female activity, and males remain highly active later into autumn.

2 What factors contribute to sex differences in activity? (Figure 3A-C, Models 3-5)

2.1 Model 3, Figure 3A: First, what are the effects of temperature and sex on nightly activity?

# response: (1 = detected | 0 = undetected) 
m3 = glmer(detect2~bat.tavg*sex + (1|site.wy2) + (1|pit_id), family = binomial(),
            control=glmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)),
            data = subset(df.m3)); summary(m3)

Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) ['glmerMod']
 Family: binomial  ( logit )
Formula: detect2 ~ bat.tavg * sex + (1 | site.wy2) + (1 | pit_id)
   Data: subset(df.m3)
Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))

     AIC      BIC   logLik deviance df.resid 
 22769.1  22816.7 -11378.5  22757.1    20773 

Scaled residuals: 
    Min      1Q  Median      3Q     Max 
-2.1133 -0.5959 -0.4535  0.7844  4.1081 

Random effects:
 Groups   Name        Variance Std.Dev.
 pit_id   (Intercept) 0.6536   0.8085  
 site.wy2 (Intercept) 0.2089   0.4571  
Number of obs: 20779, groups:  pit_id, 544; site.wy2, 9

Fixed effects:
               Estimate Std. Error z value Pr(>|z|)    
(Intercept)   -0.687705   0.217297  -3.165  0.00155 ** 
bat.tavg       0.001265   0.007507   0.169  0.86618    
sexF          -3.759775   0.627498  -5.992 2.08e-09 ***
bat.tavg:sexF  0.150806   0.030333   4.972 6.64e-07 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Correlation of Fixed Effects:
            (Intr) bt.tvg sexF  
bat.tavg    -0.662              
sexF        -0.169  0.233       
bat.tvg:sxF  0.160 -0.241 -0.980

# view temperature slopes for each sex using emtrends
kable(m3.emt)

sex	bat.tavg.trend	SE	df	asymp.LCL	asymp.UCL
M	0.0012651	0.0075070	Inf	-0.0134483	0.0159784
F	0.1520707	0.0294381	Inf	0.0943731	0.2097683

• Result: Females are less active throughout their swarm period generally, but female activity increases with temperature. Male activity is less influenced by temperature.

2.2 Model 4, Figure 3B: Then, what are the effects of the length of the active period and sex on the probability of last detection?

# response: 1 = last day active | 0 = detected on subsequent nights
m4 = glmer(last.det2~isa*sex + (1|site.wy2) + (1|pit_id), family=binomial(),
           control=glmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)),
           data=subset(df.m4));summary(m4)

Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) ['glmerMod']
 Family: binomial  ( logit )
Formula: last.det2 ~ isa * sex + (1 | site.wy2) + (1 | pit_id)
   Data: subset(df.m4)
Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))

     AIC      BIC   logLik deviance df.resid 
  2800.5   2839.4  -1394.3   2788.5     4787 

Scaled residuals: 
   Min     1Q Median     3Q    Max 
-1.665 -0.300 -0.142 -0.043 74.209 

Random effects:
 Groups   Name        Variance Std.Dev.
 pit_id   (Intercept) 5.469    2.339   
 site.wy2 (Intercept) 1.602    1.266   
Number of obs: 4793, groups:  pit_id, 433; site.wy2, 7

Fixed effects:
            Estimate Std. Error z value Pr(>|z|)    
(Intercept) -4.29912    0.70400  -6.107 1.02e-09 ***
isa          0.70560    0.07336   9.618  < 2e-16 ***
sexM        -1.37247    0.52423  -2.618  0.00884 ** 
isa:sexM    -0.36702    0.06682  -5.492 3.97e-08 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Correlation of Fixed Effects:
         (Intr) isa    sexM  
isa      -0.535              
sexM     -0.603  0.501       
isa:sexM  0.507 -0.950 -0.622

• Result: Regardless of arrival date, females spend fewer days swarming than males before ending activity.

2.3 Model 5, Figure 3C: Last, what are the effects of temperature and sex on the probability of last detection?

# response: 1 = last day active | 0 = detected on subsequent nights
m5 = glmer(last.det2~bat.tavg*sex + (1|pit_id) + (1|site.wy2), family = binomial(),
            control=glmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)),
            data = subset(df.m5)); summary(m5)

Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) ['glmerMod']
 Family: binomial  ( logit )
Formula: last.det2 ~ bat.tavg * sex + (1 | pit_id) + (1 | site.wy2)
   Data: subset(df.m5)
Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))

     AIC      BIC   logLik deviance df.resid 
  2647.9   2687.9  -1318.0   2635.9     5780 

Scaled residuals: 
    Min      1Q  Median      3Q     Max 
-1.1904 -0.2993 -0.2391 -0.1375  9.1936 

Random effects:
 Groups   Name        Variance Std.Dev.
 pit_id   (Intercept) 0.01401  0.1184  
 site.wy2 (Intercept) 0.51982  0.7210  
Number of obs: 5786, groups:  pit_id, 542; site.wy2, 9

Fixed effects:
              Estimate Std. Error z value Pr(>|z|)    
(Intercept)    0.16202    0.56715   0.286  0.77512    
bat.tavg      -0.14453    0.02669  -5.416 6.11e-08 ***
sexF           4.67270    1.54263   3.029  0.00245 ** 
bat.tavg:sexF -0.16407    0.07701  -2.131  0.03313 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Correlation of Fixed Effects:
            (Intr) bt.tvg sexF  
bat.tavg    -0.895              
sexF        -0.284  0.314       
bat.tvg:sxF  0.286 -0.321 -0.994

• Result: Females will end activity on warmer nights compared to males that remain active even as temperatures decrease.

2.4 Plot Figure 3. Contribution of temperature and length of swarm on sex-specific activity (Models 3-5)

print(p.fig3)

• Key findings: Female activity is generally reduced compared to males. Females limit nightly activity to occur at warmer temperatures, and females also end activity fewer days after arriving at the hibernacula and at higher temperatures.

3 How does mating phenology correspond to seasonal disease dynamics? (Figure 4A-B, Models 6-8)

3.1 Model 6: First, do pathogen loads differ between sexes during the autumn mating (active) season?

m6 = lmer(lgdL2~sex + (1|site.wy2), 
          data = df.m6); summary(m6) 

Linear mixed model fit by REML ['lmerMod']
Formula: lgdL2 ~ sex + (1 | site.wy2)
   Data: df.m6

REML criterion at convergence: 314.2

Scaled residuals: 
    Min      1Q  Median      3Q     Max 
-1.8127 -0.5967 -0.1576  0.5203  3.5268 

Random effects:
 Groups   Name        Variance Std.Dev.
 site.wy2 (Intercept) 0.2147   0.4634  
 Residual             0.3193   0.5650  
Number of obs: 172, groups:  site.wy2, 9

Fixed effects:
            Estimate Std. Error t value
(Intercept) -5.35210    0.17072 -31.350
sexF        -0.01397    0.09788  -0.143

Correlation of Fixed Effects:
     (Intr)
sexF -0.159

• Result: No clear infection bias during autumn mating.

3.2 Model 7, Figure 4A: Then, how do pathogen loads change seasonally by sex?

m7 = lmer(lgdL2~sex*pdate2 + (1|site.wy2), 
          data = df2); summary(m7)

Linear mixed model fit by REML ['lmerMod']
Formula: lgdL2 ~ sex * pdate2 + (1 | site.wy2)
   Data: df2

REML criterion at convergence: 822

Scaled residuals: 
    Min      1Q  Median      3Q     Max 
-2.0626 -0.6530 -0.0952  0.6261  3.5533 

Random effects:
 Groups   Name        Variance Std.Dev.
 site.wy2 (Intercept) 0.2592   0.5091  
 Residual             0.5722   0.7564  
Number of obs: 343, groups:  site.wy2, 13

Fixed effects:
             Estimate Std. Error t value
(Intercept) -11.90983    0.69417 -17.157
sexM          3.06440    0.76088   4.027
pdate2        0.74914    0.06673  11.227
sexM:pdate2  -0.34035    0.07560  -4.502

Correlation of Fixed Effects:
            (Intr) sexM   pdate2
sexM        -0.754              
pdate2      -0.971  0.776       
sexM:pdate2  0.748 -0.992 -0.783

• Result: Females develop more severe infections by early hibernation.

3.3 Model 8, Figure 4B: Last, does the end date of autumn activity influence infections in early hibernation?

m8 = lmer(lgdL2~med.maxdate + (1|site.wy2), 
           control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)),
           data = df.m8); summary(m8)

Linear mixed model fit by REML ['lmerMod']
Formula: lgdL2 ~ med.maxdate + (1 | site.wy2)
   Data: df.m8
Control: lmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))

REML criterion at convergence: 299.9

Scaled residuals: 
    Min      1Q  Median      3Q     Max 
-2.1009 -0.5816 -0.0953  0.6783  3.2574 

Random effects:
 Groups   Name        Variance Std.Dev.
 site.wy2 (Intercept) 0.6249   0.7905  
 Residual             0.6642   0.8150  
Number of obs: 114, groups:  site.wy2, 10

Fixed effects:
            Estimate Std. Error t value
(Intercept)  11.1252     3.9794   2.796
med.maxdate  -1.5740     0.4071  -3.867

Correlation of Fixed Effects:
            (Intr)
med.maxdate -0.998

3.4 Plot Figure 4. Seasonal sex-biased infection (Figure 4A-B; Models 7-8)

print(p.fig4)

• Key findings:
  • Female-biased infection only arises following the autumn mating period.
  • Infections are more severe on bats associated with sites where bats ended activity earlier.

4 Supplemental analyses and figures

4.1 Supp Figure 1: Visualize complete dataset

print(p.sf1)

• Generally, males are more active than females throughout autumn mating, fully encompassing female swarm activity.

4.2 Use balanced dataset to support sex-specific effects

4.2.1 Supp Model 1, Supp Fig 2A: Test temperature-dependence of bat activity

sm1 = glmer(detect2~bat.tavg*sex + (1|site.wy2) + (1|pit_id), family = binomial(),
             control=glmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)),
             data = subset(df1.tru)); summary(sm1)

Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) ['glmerMod']
 Family: binomial  ( logit )
Formula: detect2 ~ bat.tavg * sex + (1 | site.wy2) + (1 | pit_id)
   Data: subset(df1.tru)
Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))

     AIC      BIC   logLik deviance df.resid 
  4354.4   4392.6  -2171.2   4342.4     4266 

Scaled residuals: 
    Min      1Q  Median      3Q     Max 
-1.5418 -0.5403 -0.4081 -0.2618  4.1241 

Random effects:
 Groups   Name        Variance Std.Dev.
 pit_id   (Intercept) 0.5480   0.7402  
 site.wy2 (Intercept) 0.0166   0.1288  
Number of obs: 4272, groups:  pit_id, 178; site.wy2, 8

Fixed effects:
              Estimate Std. Error z value Pr(>|z|)    
(Intercept)   -1.19533    0.43121  -2.772  0.00557 ** 
bat.tavg       0.02066    0.02126   0.972  0.33107    
sexF          -3.17607    0.74400  -4.269 1.96e-05 ***
bat.tavg:sexF  0.11847    0.03614   3.279  0.00104 ** 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Correlation of Fixed Effects:
            (Intr) bt.tvg sexF  
bat.tavg    -0.964              
sexF        -0.573  0.565       
bat.tvg:sxF  0.563 -0.582 -0.980

#view temperature trends by sex
kable(sm1.emt)

sex	bat.tavg.trend	SE	df	asymp.LCL	asymp.UCL
M	0.0206591	0.0212553	Inf	-0.0210005	0.0623187
F	0.1391326	0.0293756	Inf	0.0815574	0.1967078

4.2.2 Supp Model 2, Supp Fig 2B: Differences in nightly temperature by sex and activity

sm2 = lmer(bat.tavg~detect2*sex + (1|pit_id) + (1|site.wy2), 
           data = subset(df1.tru)); summary(sm2)

Linear mixed model fit by REML ['lmerMod']
Formula: bat.tavg ~ detect2 * sex + (1 | pit_id) + (1 | site.wy2)
   Data: subset(df1.tru)

REML criterion at convergence: 19136

Scaled residuals: 
    Min      1Q  Median      3Q     Max 
-3.1646 -0.6908 -0.0348  0.7159  3.1546 

Random effects:
 Groups   Name        Variance Std.Dev.
 pit_id   (Intercept) 0.3333   0.5773  
 site.wy2 (Intercept) 1.3436   1.1591  
 Residual             4.9435   2.2234  
Number of obs: 4272, groups:  pit_id, 178; site.wy2, 8

Fixed effects:
              Estimate Std. Error t value
(Intercept)   19.35260    0.42333  45.715
detect21       0.13314    0.10616   1.254
sexF          -0.03922    0.12987  -0.302
detect21:sexF  0.48511    0.17724   2.737

Correlation of Fixed Effects:
            (Intr) dtct21 sexF  
detect21    -0.086              
sexF        -0.138  0.264       
detct21:sxF  0.051 -0.600 -0.322

4.2.3 Plot Supp Figure 2. Effects of temperature with balanced observations between sexes

print(p.sf2)

• Results consistent with Model 3 in Figure 3A, using the balanced dataset between sexes:
  • Females increase nightly activity with temperature compared to males whose activity is relatively unaffected by temperature.
  • The difference between mean temperature on nights when bats were active compared to night when bats were undetected was greater for females than males. Females concentrate their activity to warmer nights compared to males that do not.

4.3 Does body condition affect nightly activity (Supp Figure 3; Supp Model 3)?

4.3.1 Test whether temperature and mass differently influence activity between sexes

sm3 <- glmmTMB(detect2~sex*mass*temp + (1|pit_id),
               family = binomial(),
               data = subset(df.sm3)); summary(sm3)

 Family: binomial  ( logit )
Formula:          detect2 ~ sex * mass * temp + (1 | pit_id)
Data: subset(df.sm3)

     AIC      BIC   logLik deviance df.resid 
  5146.4   5203.6  -2564.2   5128.4     4249 

Random effects:

Conditional model:
 Groups Name        Variance Std.Dev.
 pit_id (Intercept) 0.9313   0.965   
Number of obs: 4258, groups:  pit_id, 336

Conditional model:
                Estimate Std. Error z value Pr(>|z|)  
(Intercept)     -4.82732    2.28755  -2.110   0.0348 *
sexF           -19.80829    9.78416  -2.025   0.0429 *
mass             0.34114    0.25653   1.330   0.1836  
temp             0.10615    0.11711   0.906   0.3647  
sexF:mass        2.20187    1.17849   1.868   0.0617 .
sexF:temp        1.03982    0.49003   2.122   0.0338 *
mass:temp       -0.00289    0.01312  -0.220   0.8256  
sexF:mass:temp  -0.12246    0.05921  -2.068   0.0386 *
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

4.3.2 Plot Supp Figure 3: Effects of mass and temperature on nightly activity

print(p.sf3)

• Results:
  • Females with lower body condition (ie body mass) concentrated activity to warm nights, compared to females with higher body condition.
  • The temperatures at which males were active was not affected by their body condition.
  • Females adjust nightly activity according to their body condition but males do not, suggesting the sexes are likely budgeting their energy differently.

4.4 Additional analysis showing association between between mating phenology and disease (Supp Figure 4; Supp Figure 4)

4.4.1 Test whether median end date of autumn activity at a site influenced mean pathogen loads in early hibernation

sm4 = lmer(mean.lgdL~med.maxdate + (1|site.wy2), data = subset(df2.sum)); summary(sm4)

Linear mixed model fit by REML ['lmerMod']
Formula: mean.lgdL ~ med.maxdate + (1 | site.wy2)
   Data: subset(df2.sum)

REML criterion at convergence: 31.1

Scaled residuals: 
     Min       1Q   Median       3Q      Max 
-0.82223 -0.53563 -0.06433  0.47551  1.18360 

Random effects:
 Groups   Name        Variance Std.Dev.
 site.wy2 (Intercept) 0.5788   0.7608  
 Residual             0.1053   0.3244  
Number of obs: 16, groups:  site.wy2, 10

Fixed effects:
            Estimate Std. Error t value
(Intercept)  11.0041     3.1930   3.446
med.maxdate  -1.5615     0.3285  -4.753

Correlation of Fixed Effects:
            (Intr)
med.maxdate -0.997

4.4.2 Plot Supp Figure 4. Effect of median end date of autumn activity on mean early hibernation infections at site.

print(p.sf4)

• Results: Consistent with Figure 4B, infection severity in early hibernation is more severe at sites where autumn activity ended sooner.

4.5 How many nights of activity do individuals accumulate throughout the length of their active period?

4.5.1 Plot Supp Figure 5: Visualize cumulative number of active nights by sex

print(p.sf5)

4.5.2 Calculate average number of nights detected among individuals of each sex

kable(n.dates.sum)

sex	mean_n.dates
F	3.810526
M	13.947471

• Males continue to be active without any clear threshold in the number of dates active, whereas females cease activity after an average of 4 nights.

4.6 Model comparisons and validation

4.6.1 Compare activity period models as additive or interactive with site-year and sex as fixed effects (Supp Table 2)

First, begin date of activity models

m1b.add = lmer(mindate2~site.wy2 + sex + (1|pit_id), data = df.m1)
m1b.int = lmer(mindate2~site.wy2 * sex + (1|pit_id), data = df.m1)

• Compare using AIC

kable(AIC(m1b.add,m1b.int))

	df	AIC
m1b.add	10	1142.960
m1b.int	16	1160.287

Then, end date of activity models

m2b.add = lmer(maxdate2~site.wy2 + sex + (1|pit_id), data = df.m2)
m2b.int = lmer(maxdate2~site.wy2 * sex + (1|pit_id), data = df.m2)

• Compre using AIC

kable(AIC(m2b.add,m2b.int))

	df	AIC
m2b.add	13	1692.986
m2b.int	20	1711.309

• Key takeaway:
  • Sex is more supported as an additive rather than interactive effect in models including site-year as fixed effects.
  • The dates that activity started and ended varied among sites and years, but females consistently were later to start and earlier to end.

4.6.2 View results of null model comparisons and AUC estimates from k-fold cross validation of binomial models

kable(nullcomps_full)

	df	AIC
m1.null	4	1134.2857
m1	5	1129.1469
m2.null	4	1715.0409
m2	5	1682.9043
m3.null	3	22821.9059
m3	6	22769.0835
m4.null	2	3623.1434
m4	6	2800.5278
m5.null	2	2741.8440
m5	6	2647.9294
m7.null	3	966.0277
m7	6	833.9688
m8.null	3	319.9266
m8	4	307.8518

• All AICs of reported models are >2 scores below the null models (structured with response~1), indicating improvement over the null.

4.6.3 View AUC scores derived from k-fold cross validation of the logistic models reported in the main results.

kable(kfold_full)

Model	AUC
m3	70.54%
m4	75.94%
m5	83.09%

• Results indicate that >70%, 75%, and 83% of our test data was successfully predicted by the training models of Model 3, Model 4, and Model 5, respectively.

4.7 View sampling distributions by site, year and sex where applicable by model

ST3A. Model 1 Sample summary: Start dates of activity
Year	Females	Males
BA IT
2021	11	60
2022	16	134
MA N
2021	13	33
2022	10	72
NE MI
2021	2	36
2022	17	88
2023	26	92

ST3B. Model 2 Sample summary: End dates of activity
Year	Females	Males
BA IT
2020	4	63
2021	12	102
2022	17	139
MA N
2020		10
2021	18	73
2022	16	98
NE MI
2020		20
2021	7	73
2022	24	107
2023	35	130

ST3C. Model 3 Sample summary: Prob of nightly activity
Year	Females	Males
BA IT
2020	4	63
2021	12	101
2022	17	139
MA N
2020		10
2021	18	71
2022	16	100
NE MI
2020		20
2021	7	73
2022	24	107

ST3D. Model 4 Sample summary: Prob of last detection
Year	Females	Males
BA IT
2021	11	60
2022	16	134
MA N
2021	13	33
2022	10	73
NE MI
2021	2	36
2022	17	88
2023	26	92

ST3E. Model 5 Sample summary: Prob of last detection
Year	Females	Males
BA IT
2020	4	63
2021	12	97
2022	17	139
MA N
2020		10
2021	18	67
2022	16	100
NE MI
2020		20
2021	6	72
2022	24	107

ST3F. Model 7 Sample summary: Seasonal infections
Year	Season	Females	Males
BA IT
2019	Early hiber	10	9
2020	Autumn mating	26	66
2020	Early hiber	5	12
2021	Autumn mating		3
2021	Early hiber	1	10
2022	Autumn mating	11	38
2022	Early hiber	3	13
MA N
2019	Early hiber	4	15
2020	Autumn mating	12	28
2020	Early hiber	5	11
2021	Autumn mating	14	12
2021	Early hiber	2	7
2022	Autumn mating	22	48
2022	Early hiber	4	9
NE MI
2019	Early hiber	5	13
2020	Autumn mating	5	37
2020	Early hiber	1	11
2021	Autumn mating		6
2021	Early hiber	5	8
2022	Autumn mating	23	34
2022	Early hiber	10	6
2023	Early hiber		20

ST3G. Model 8 Sample summary: Phenology-dependent early hibernation infections
Year	N
BA IT
2020	16
2021	8
2022	10
MA N
2020	8
2021	9
2022	10
NE MI
2020	7
2021	12
2022	15
2023	19